Some content management systems facilitate sharing content items and collections amongst users through the use of distinct namespaces. In current approaches, software running on a client device identifies shared collections contained within a synchronized collection. The shared collections are then synchronized with their respective namespaces while the remainder of the synchronized collection is synchronized with another namespace. In current approaches, the software running on the client device actively manages the assignments between collections and namespaces. However, managing multiple namespaces within a client software environment introduces development and processing complexity. Namespace management within the client software environment can also introduce correctness issues and violate the principle of only having a “single source of truth.” What is needed is an approach that more intelligently manages namespaces.